Setting levels to tape machine: understanding where saturation/distortion might begin

[cashhewn]

Hi,

Regarding its specs, the manual for the Tascam 22-2 states:
“Line Input:
Input impedance: 50k ohms, unbalanced
Maximum source impedance: 10k ohms or less
Nominal input level: -10 dBv (0.3 V)
Minimum input level: -20 dBv (100 mV)”

My brain seems to understand my studio gear best in dBu, so if -10dBv is equal to -7.79dBu, then for me to hit the tape machine input at its “nominal level” then the RMS of my signal needs to be hovering around -8dBu/-12VU/.308V.

Besides monitoring off the tape head and guessing where saturation/distortion begins, is there any way to estimate (even roughly) where say, the transient of a kick or snare might begin?

For example I’m able to know via the DAW and my converters that if I have a kick/snare peaking at -10 dBfs which is +9dBu out of my D/A, that’s 17VU over the “nominal input” of the deck (and iirc correctly saturates/distorts heavily), so:

Is there some calculable best guess of VU/dBu level above the tape machine’s nominal input where I begin to leave “clean reproduction” realm and enter the saturation zone?

Many thanks.

[Hugh Robjohns]

...for me to hit the tape machine input at its “nominal level” then the RMS of my signal needs to be hovering around -8dBu/-12VU/.308V.

Assuming the tape machine is aligned to spec, and your meters are calibrated 0VU=+4dBu, then yes.

Besides monitoring off the tape head and guessing where saturation/distortion begins...

You don't have to guess — you can hear it in real-time. That's the whole point of listening off-tape!

...Is there some calculable best guess of VU/dBu level above the tape machine’s nominal input where I begin to leave “clean reproduction” realm and enter the saturation zone?

Yes....kind of... but it really depends (mostly on the brand of tape and the machine line up.

I dont know about the Tascam, but typically tape is lined up to give 3% THD at 0VU, and distortion (or saturation if you prefer) builds above that.

Typically, when we wanted saturation we'd run at +6dB or even +9dB above 0VU*... but since VU meters normally hit the ends top just above +3VU it's really guesswork and experience.

*...Driving the tape that hard assumes that (a) the tape can take it and (b) the tape recorders heads and amps have sufficient headroom too.

[cashhewn]

Many thanks Hugh, this is very helpful and much appreciated.

[James Perrett]

If you follow the proper line up process which will be described in any decent tape machine manual then all these questions will be answered. The line up process and your choice of tape gives you control over the level where you want saturation to start to become apparent.

[SafeandSound Mastering]

Slightly less numbers driven and very sensible to do. Make a test recording and find out.

Hit 0VU or slightly under with the below mentioned caveat and listen back to the recording.

An important consideration is knowing that high frequency, high transient energy sources were generally recorded lower than 0VU (which reduces SNR of course but reduces gross distortion risk). The VU meter could not track the transient dynamic swing as the rise time is not fast enough. As such did not read the true transient levels of, for example, overheads - hi hats and cymbals.

As far as distortion goes, also worth knowing tape is immediately much more distorted than correct gain structure digital recording (though it may not be obviously noticeable, but it is part of the sound character nonetheless.). It might be in single figure THD even at nominal recording levels before going anywhere near the red on a middle of the road machine/tape formulation.

[Wonks]

Though here, without any recalibration, 0VU for the domestic level line inputs on the recorder = -12VU, as cashewn previously mentioned. But at least that gives the ability to measure higher signal levels than -12VU going to the tape machine and the resulting distortion levels.

I'd suggest using a software VU meter like the Klanghelm VUMT, where you can easily set the dBFS calibration level for 0VU (here -27dBFS would seem to be appropriate) and use meter displays that behave like analogue VU meters. Doing that does limit your ability to accurately see how much the signal is going over your nominal 0VU, but you can obviously have several meter instances going with different calibration levels to cover that.

[Hugh Robjohns]

As I said earlier — and I think Safe&Sound Mastering is saying the same thing — you can't easily calculate the boundary between clean and saturated with analogue tape.

Partly, thats because its a progressive change, not a step function, and also because any numerical assessment would need a lot more information than you've given (or have available)!

So, you have to judge it empirically, using your ears, and build your experience with that particular tape machine with its current alignment running that particular tape formulation.

So, experiment with higher and lower levels relative to your normal working practice and listen to what comes back 60ms later (or however long the delay between rec and rep heads on your machine). Note how the transients and sustained sounds are affected at different fader and/or meter levels. How the low frequency sounds change relative to high frequencies at different levels.

Normally, in professional circles, a machine is calibrated for a specific tape type and only that tape type is used on that machine... but you may decide to use alternative tape types without recalibration the machine, in which case a the characteristics previously learned will change... which may be used as a beneficial musical choice and effect, of course.

This is all part of the skill (or art if you prefer) of analogue recording.... because every tape machine, tape type, and alignment will inherently sound and behave differently, and respond differently as you push levels by varying amounts!

[SafeandSound Mastering]

It sprang to mind without knowing the tape machine in question, if you have phono inputs that might be expecting -10dBV and you may have +4dBu reference outputs from for example an audio interface.

It is a quite a difference in voltage level (1.23V vs 0.316V RMS). This is about 12dB difference. (double and double again gets you close so 6dB + 6dB for voltage)

This might be a little helpful:

https://sengpielaudio.com/calculator-db-volt.htm

It could be possible (before tape saturation concerns) that you would overload the electronics in the input stages (with +4dBu outputs driving -10dBV inputs) before it gets to tape and then you could be potentially just padding down an already electronically clipped signal onto tape by using the input gain to tape.

As wonks mentioned you could need what would typically appear to be a fairly low output level in dBFS to have correct nominal signal (if you do have the disparity between ref levels between output and inputs).

I did not mention this before, the above is worth considering.

And probably already covered as I see posts above that I did not fully read before going into full on type mode.

Some audio interfaces like RME allow you to switch the output levels between reference levels.

You should have all you need to suss it now.

[Hugh Robjohns]

...or make life easy and use a pro/consumer conversion box that translates between the balanced +4dBu world and unbalanced -10dBV world, in both directions — allowing the use of pro desk or interface normally without having to do the mental arithmetical to understand meter levels.

[cashhewn]

...or make life easy and use a pro/consumer conversion box that translates between the balanced +4dBu world and unbalanced -10dBV world, in both directions — allowing the use of pro desk or interface normally without having to do the mental arithmetical to understand meter levels.

I appreciate all of the responses, James, Wonks, Safe and Sound.

Hugh, you’ve mentioned the Art Cleanbox Pro for this purpose before. To speak frankly, I got confused and prevented myself from pulling the trigger on a pair because for one, I feel like the signal flow diagram is flawed in the manual (see link at end of post, at the far right or top of the diagram it shows unbalanced outputs on XLR connections that should clearly be balanced ones, no?), and then two, those pots for “stereo output level” on the front, where should those be set to simply maintain unity +4dBu/-10dBv conversion in both directions?

If Hugh endorses them then I’d guarantee that they do what they say on the tin, and I may well purchase a pair, but couldn’t I also attenuate and boost by just about 12dB in both directions, (the difference between +4dBu/-10dBv I believe) if I could figure out existing kit to do exactly 12dB of atten/boost?

https://artproaudio.com/imgs/file/Manu ... l_V1_3.pdf

[Arpangel]

...or make life easy and use a pro/consumer conversion box that translates between the balanced +4dBu world and unbalanced -10dBV world, in both directions — allowing the use of pro desk or interface normally without having to do the mental arithmetical to understand meter levels.

I appreciate all of the responses, James, Wonks, Safe and Sound.

Hugh, you’ve mentioned the Art Cleanbox Pro for this purpose before. To speak frankly, I got confused and prevented myself from pulling the trigger on a pair because for one, I feel like the signal flow diagram is flawed in the manual (see link at end of post, at the far right or top of the diagram it shows unbalanced outputs on XLR connections that should clearly be balanced ones, no?), and then two, those pots for “stereo output level” on the front, where should those be set to simply maintain unity +4dBu/-10dBv conversion in both directions?

If Hugh endorses them then I’d guarantee that they do what they say on the tin, and I may well purchase a pair, but couldn’t I also attenuate and boost by just about 12dB in both directions, (the difference between +4dBu/-10dBv I believe) if I could figure out existing kit to do exactly 12dB of atten/boost?

https://artproaudio.com/imgs/file/Manu ... l_V1_3.pdf

Don’t worry about the Clean Box, it’s fine, I use them for various things and they just disappear, do the job. If you can’t hear it don’t worry about it.
Regarding tape saturation, it's a creative tool, you have to use your ears, and it varies so much between machines, I've had machines that have been hard to overload, and those that overload at the drop of a hat, you can play around with settings, but honestly, I just accept it for what it is, head wear, alignment, tape brands and types, just play around until you get something you like the sound of, I've never aligned a tape machine in my life, wouldn’t know how to, but I’ve owned loads of them, and enjoyed working with them, they are great fun.

[Hugh Robjohns]

...I feel like the signal flow diagram is flawed in the manual...

It's very obviously a typo in the output label!

... those pots for “stereo output level” on the front, where should those be set to simply maintain unity +4dBu/-10dBv conversion in both directions?

Around 12 noon... but they make it very easy to calibrate your system levels precisely.

Play steady tone from your mixer at 0VU to the tape recorder (monitoring the input), and adjust the CleanBox Pro unbalanced stereo output level control to get 0VU on the recorder meter.

Record 30 seconds of that tone and replay it. If the machine is lined up correctly it should read 0VU on the machine's meters during playback. Adjust the CleanBox Pro balanced stereo output level control to get 0VU on the desk meter. Calibration done!

(If the machine replay is not 0VU adjust the level control to match the reading, whatever it is, on the desk.)

... couldn’t I also attenuate and boost by just about 12dB in both directions, (the difference between +4dBu/-10dBv I believe) if I could figure out existing kit to do exactly 12dB of atten/boost?

Strictly, (just for the pedants) it's 11.8dB, but 12dB is close enough for me. It's not 14dB anyway, which is a common mistake people make.

You could fudge ways to attenuate and boost the level in your desk and tape machine, yes.... but this way avoids bodge leads and adapters, and you don't have to figure out anything. Everything just works at optimum signal levels with no compromises to noise floor or headroom performances.

Plenty of other companies make pro/consumer converter boxes too, besides ART, of course.

[cashhewn]

Arpangel: thanks, I appreciate the sentiment and agree.

It's very obviously a typo in the output label!

Fantastic, at least I’ve learned to read a simple block diagram…

Play a steady tone from your mixer at 0VU to the tape recorder (monitoring the input), and adjust the CleanBox Pro unbalanced stereo output level control to get 0VU on the recorder meter.

Record 30 seconds of that tone and replay it. If the machine is lined up correctly it should read 0VU on the machine's meters during playback. Adjust the CleanBox Pro balanced stereo output level control to get 0VU on the desk meter. Calibration done!

(If the machine replay is not 0VU adjust the level control to match the reading, whatever it is, on the desk.)

Many thanks for this detailed step-by-step information.

Just as it seems we’re putting this to bed, I’m going to mention that I went through my notes and the deck was last setup about four years ago for Ampex 641, for whatever that’s worth…

[Hugh Robjohns]

As I recall, 641 was long-play (1mil thick) tape intended for consumer machines dating back to the early 60s.

It wasn't very common over this side of the Atlantic, but I think it was considered a very good formulation in its day, and was still being made in the Quantegy era.

It won't take the high flux levels that the last generations of tape recorders could manage, and it wasn't a particularly quiet tape, though.

And being Ampex it's likely to suffer sticky-shed syndrome if not stored in appropriate conditions.

[James Perrett]

I’m going to mention that I went through my notes and the deck was last setup about four years ago for Ampex 641, for whatever that’s worth…

I hope you aren't going to try using any Ampex tape on the machine nowadays. You'll end up clogging up the machine and, if you aren't careful, possibly damaging the motors. Even well stored Ampex tape is going to be sticky now.

RTM LPR35 is probably the nearest you can buy nowadays and should work well.

[cashhewn]

I hope you aren't going to try using any Ampex tape on the machine nowadays. You'll end up clogging up the machine and, if you aren't careful, possibly damaging the motors. Even well stored Ampex tape is going to be sticky now.

RTM LPR35 is probably the nearest you can buy nowadays and should work well.

Thanks James, I guess I’ve been lucky with my 641 stock, no issues here.

Pardon my tape stock ignorance, I’m in the learning stage, but as far as the RTM LPR35 tape goes, if it’s a “standard bias” tape and it’s the “same formula as the high bias studio tape SM911” which is +6, what makes it standard bias, the increased thickness? Or?

I agree that it appears the LPR35 seems a good spool-up-and-go replacement for the 641 without needing to bias, thanks for that suggestion. It’s a slightly more cost effective stock than the 911 and it seems I won’t need to bias it.

Next stop on the train is learning to check levels through the deck and bias…

[Tim Gillett]

As I recall, 641...

...And being Ampex it's likely to suffer sticky-shed syndrome if not stored in appropriate conditions.

641 was a good tape, made prior to the later Sticky Shed tapes.

https://reeltoreelwarehouse.com/brand_info/ampex/

The Sticky Shed tapes tend to develop their problems even when stored in good conditions. Ted Kendall's advice is basically "bake SSS tapes on sight".

[Arpangel]

I used to use Scotch tape, can’t remember the number, but I preferred the sound to Ampex.
I found 45 reels of 2 inch Ampex in a skip, I sold most of it but I had to give the money back as it was all shedding.

[James Perrett]

Thanks James, I guess I’ve been lucky with my 641 stock, no issues here.

Pardon my tape stock ignorance, I’m in the learning stage, but as far as the RTM LPR35 tape goes, if it’s a “standard bias” tape and it’s the “same formula as the high bias studio tape SM911” which is +6, what makes it standard bias, the increased thickness? Or?

I think you are confusing bias level with recording level. SM911 requires a similar bias level to most modern ferric tapes but can handle higher signal levels with lower distortion than standard tapes so the standard operating level is usually set to +6dB above the 320nWb/m standard. SM900 can handle even higher levels.

The full specs can be found at

https://cdn.shopify.com/s/files/1/0602/ ... 1732725119

If you want to use the tape sound as an effect then you might be better off using older tapes as your machine will drive them into saturation more easily.

[Hugh Robjohns]

For the avoidance of confusion, the bias signal is a very high frequency (typically 50-150kHz) tone mixed in with the wanted audio which essentially linearises the recording process, reducing distortion.

Each different tape formulation requires a different amount of bias for optimum performance, and the bias level also varies with tape speed. Incorrect bias degrades the high-frequency response, the noise floor, and the amount of distortion (at all signal levels).

The recorded level on tape is referenced to a particular flux level, and the machine's record and replay amps are aligned such that a signal at the chosen reference flux indicates 0VU on the meters.

Special 'alignment tapes' are available (at enormous expense) which are recorded at a specific reference flux level, and with tones at different frequencies to allow the replay EQ to be adjusted to the correct standard, thus ensuring correct replay frequency response. (The record side is aligned to match the replay! Most alignment tapes these days have a flux level of 320nWb/m, but other ref levels are available.

Early machines typically used 185nWb/m as the reference flux level, but later formulations could take 200, 250, 320nWb/m or more — and as a result these 'high output' tapes effectively allowed greater dynamic range and lower noise floors — but only if the machine's tape heads and amplifiers were designed to handle these elevated flux levels.

During my time at the BBC the tape alignments gradually crept higher as newer tape formulations were adopted.

The final generation of stereo tape machines through the 1990s were lined up at a reference level of 400nWb/m (+2dB above 320) with peaks to 1000nWb/m.

Mono machines and multitracks were aligned at 250nWb/m (-2dB below 320) if memory serves.

+6dB above 320nWb/m would be 640nWb/m.

[cashhewn]

Hugh, thanks for the thorough explanation, it’s very much appreciated.

There’s obviously quite a lot to maintaining a tape machine. So far the deck has performed really well since I bought it fully serviced around five years ago, but it seems now that test tapes are probably in my future if I’m going to keep it up (and there are very few tape machine service options any longer).

Fortunately the 22-2 manual is very good. If I’m reading it right it seems like everything can be done with: two test tapes (for record and reproduce it seems one is DIN250nWb/m, 31.5 Hz-16 kHz, and the other is NAB, 20 kHz-31.5 Hz), a voltmeter, a signal generator, and a 10k ohm load.

The cleaning of the heads and lubricating of the rollers is much more straightforward!

What does one need to demagnetize the tape path?

[Martin Walker]

What does one need to demagnetize the tape path?

Something like this:

https://www.amazon.co.uk/Demagnetiser-T ... 50&sr=8-10

[Hugh Robjohns]

... it seems now that test tapes are probably in my future if I’m going to keep it up ...

Good idea if you're going to service the machine yourself. But make sure you're sitting down when checking the price!

MRL are a well-known and trusted source of test tapes, and their website has a wealth of helpful information:

https://www.mrltapes.com/

If I’m reading it right it seems like everything can be done with: two test tapes (for record and reproduce it seems one is DIN250nWb/m, 31.5 Hz-16 kHz, and the other is NAB, 20 kHz-31.5 Hz), a voltmeter, a signal generator, and a 10k ohm load.

By voltmeter, they probably mean an accurate audio level meter, as opposed to a multimeter!

You normally need a test tape for each speed the machine runs at, recorded to whatever recording EQ standard the machine uses — NAB, IEC(DIN), or AES. You can use offset levels to line up a NAB machine with an IEC tape, or vice versa, but I wouldn't recommend it.

Alignment tapes are available with different features, but most will include 1kHz tone at the reference level, 10kHz tone for azimuth alignment, and a sequence of frequency tones from 31Hz to 20kHz to check and tweak the replay EQ — the relative levels of these being different for NAB or IEC etc.

The record side of the machine is setup using your preferred brand of (new) tape, tweaking levels, azimuth and record EQ to match the calibrated replay electronics.

What does one need to demagnetize the tape path?

Skill and a demagnetiser!

I recommend the Han-d-Mag model — available in 115 and 230V models — from RB Annis instruments. It's not cheap, but it is powerful and very effective. I've found some inexpensive models dont have sufficient power to demagnetise properly in some cases!

https://rbannis.com/han-d-mags/

RB Annis offer the demagger on its own, or in an excellent Han-D-Kit with a magnetometer (which is very helpful in quantifying the level of magnetism before and after demagnetising) and a couple of steel bars which you can use to practice and hone your demag skills before going anywhere close to the machine!

This is important because if you don't do it properly its very easy to end up magnetising the machine instead of demagnetising it!

[cashhewn]

Martin and Hugh, thank you for the links.

And Hugh, thank you so much for the detailed and thorough information, it is much appreciated!

Demagnetization: now I’m nervous…

Fortunately it turns out that I know someone in my area that maintained tape machines at one of the big houses in LA for a large part of the 80s/90s. I’m going to look into getting the tapes and kit over the coming months and see about some consultation/tutelage from this person.

Thanks again for your help!

[Hugh Robjohns]